Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for selecting data content to be pushed to a terminal, comprising: acquiring a user identifier, and acquiring a characteristic value, corresponding to the user identifier, in a preset user attribute type; acquiring data content, and searching for a decision tree object corresponding to the data content, wherein a tree node of the decision tree object comprises a branch node and a leaf node, the branch node has a one-to-one correspondence with the user attribute type, and a characteristic threshold of each characteristic section in the user attribute type is stored in the branch node corresponding to the user attribute type, a sub node of the branch node has a one-to-one correspondence with the characteristic threshold, and the number of clicks and the number of pushes corresponding to a characteristic threshold corresponding to the leaf node are stored in the leaf node; locating a leaf node corresponding to the user identifier in the decision tree object based on the characteristic value, corresponding to the user identifier, in the preset user attribute type, wherein the characteristic value matches with a characteristic threshold corresponding to each tree node on a path from a root node of the decision tree object to the located leaf node; and acquiring the number of clicks and the number of pushes stored in the located leaf node, generating a selection reference value based on the number of clicks and the number of pushes, and selecting, based on the selection reference value, data content to be pushed to a terminal corresponding to the user identifier, wherein the method is performed by a processor.
This invention relates to data content selection for user terminals and addresses the problem of efficiently and intelligently pushing relevant data to users. The method involves acquiring a unique user identifier and a corresponding characteristic value from a predefined user attribute type. Data content is then obtained, and a decision tree object associated with this data content is searched. The decision tree object is structured with branch nodes and leaf nodes. Each branch node corresponds to a user attribute type and stores characteristic thresholds for different sections within that attribute type. Sub-nodes of a branch node correspond to these characteristic thresholds. Leaf nodes store the number of clicks and the number of pushes for a specific characteristic threshold. The system then locates a specific leaf node within the decision tree object that matches the user's characteristic value. This matching occurs by traversing the tree from the root node, with the user's characteristic value aligning with characteristic thresholds at each node along the path to the identified leaf node. Finally, the number of clicks and pushes stored in the located leaf node are retrieved. A selection reference value is generated based on these metrics, and this value is used to select the data content to be pushed to the user's terminal. The entire process is executed by a processor.
2. The method for selecting data content to be pushed to a terminal according to claim 1 , further comprising: receiving an uploaded browsing history, and acquiring a user identifier corresponding to the uploaded browsing history and data content corresponding to the uploaded browsing history; and acquiring a decision tree object corresponding to the data content, acquiring a characteristic value, corresponding to the user identifier, in the preset user attribute type, locating a leaf node corresponding to the user identifier in the decision tree object based on the acquired characteristic value, and increasing the number of clicks and the number of pushes stored in the located leaf node based on the uploaded browsing history.
This invention relates to personalized data content delivery systems, specifically methods for selecting and pushing relevant content to user terminals based on browsing history and user attributes. The problem addressed is the inefficiency of generic content recommendations, which often fail to account for individual user preferences and behaviors. The method involves receiving a user's browsing history, which includes records of previously accessed data content. From this history, a user identifier and the corresponding data content are extracted. A decision tree object associated with the data content is then retrieved. The system acquires a characteristic value for the user identifier from a predefined set of user attribute types (e.g., demographics, interests, or behavior patterns). Using this characteristic value, the system locates a specific leaf node in the decision tree that corresponds to the user. The leaf node stores metrics such as the number of clicks and the number of pushes for the content. Based on the uploaded browsing history, these metrics are updated to reflect the user's interaction with the content. This process helps refine future content recommendations by dynamically adjusting the decision tree structure to better match user preferences. The system ensures that content pushed to the user is more relevant, improving engagement and reducing irrelevant content delivery.
3. The method for selecting data content to be pushed to a terminal according to claim 2 , wherein increasing the number of clicks and the number of pushes stored in the located leaf node based on the uploaded browsing history further comprises: acquiring the number of clicks and the number of pushes, corresponding to the data content, in the uploaded browsing history; and acquiring a branch node on the path from the root node to the located leaf node in the decision tree object, acquiring a preset candidate user attribute type besides a user attribute type corresponding to the branch node on the path, and adding, by category, the number of clicks and the number of pushes, corresponding to the data content, acquired in the uploaded browsing history based on each characteristic section in the candidate user attribute type.
This invention relates to a method for selecting and pushing data content to a terminal based on user browsing history. The problem addressed is efficiently determining which content to push to users by analyzing their browsing behavior and updating a decision tree structure to improve content recommendations. The method involves a decision tree object with nodes representing user attributes and paths leading to leaf nodes that store content recommendations. When a user's browsing history is uploaded, the system locates the corresponding leaf node in the decision tree. The number of clicks and pushes associated with the data content in the browsing history is then acquired. These values are used to update both the leaf node and intermediate branch nodes along the path from the root to the leaf. For each branch node, the system identifies a preset candidate user attribute type (different from the attribute type of the branch node) and updates the click and push counts for each characteristic section within that candidate attribute type. This ensures that the decision tree adapts to user behavior across multiple dimensions, improving the accuracy of future content recommendations. The method dynamically refines the decision tree structure to optimize content delivery based on real-time user interactions.
4. The method for selecting data content to be pushed to a terminal according to claim 3 , further comprising after adding, by category, the number of clicks and the number of pushes, corresponding to the data content, acquired in the uploaded browsing history based on each characteristic section in the candidate user attribute type: generating an information gain corresponding to the candidate user attribute type based on the number of clicks and the number of pushes, corresponding to each characteristic section in the candidate user attribute type, stored in the located leaf node by category; searching for a candidate user attribute type having an information gain, a difference between which and an information gain of other searched user attribute type is greater than or equal to an information gain threshold; and setting the located leaf node as a branch node and generating a leaf node of the branch node based on a characteristic threshold of a characteristic section in the found candidate user attribute type, in a case that the candidate user attribute type is found.
This invention relates to a method for selecting and pushing data content to a terminal based on user attributes and browsing history. The method addresses the challenge of efficiently delivering relevant content to users by analyzing their browsing behavior and attribute data to improve content recommendation accuracy. The method involves collecting browsing history data, which includes user interactions such as clicks and content pushes. This data is categorized by user attributes, such as demographic or behavioral characteristics, and stored in a decision tree structure. Each node in the tree represents a user attribute type, and the method calculates an information gain for each attribute type based on the number of clicks and pushes associated with the content. Information gain measures the reduction in uncertainty about user preferences when considering a particular attribute. The method then identifies candidate user attribute types where the information gain significantly exceeds that of other attributes, as determined by a predefined threshold. If such an attribute is found, the current node in the decision tree is converted into a branch node, and new leaf nodes are generated based on characteristic thresholds derived from the selected attribute. This process refines the decision tree, allowing for more precise content recommendations tailored to user attributes. The method dynamically updates the tree structure to adapt to changing user behavior and preferences.
5. The method for selecting data content to be pushed to a terminal according to claim 4 , wherein generating the information gain corresponding to the candidate user attribute type based on the number of clicks and the number of pushes, corresponding to each characteristic section in the candidate user attribute type, stored in the located leaf node by category comprises: calculating an information gain of a user attribute type A of a leaf node S according to: { G ( A ) = Entropy ( S ) - ∑ v ∈ F A p ( v ) Entropy ( S v ) Entropy ( S ) = - p 1 log 2 ( p 1 ) - ( 1 - p 1 ) log 2 ( 1 - p 1 ) Entropy ( S v ) = - p 2 log 2 ( p 2 ) - ( 1 - p 2 ) log 2 ( 1 - p 2 ) wherein F A is a set of characteristic sections in the user attribute type A, v is a characteristic threshold of each characteristic section in the user attribute type A, p(v) is a distribution probability of the number of pushes in each characteristic section in the user attribute type A, S v is a set of the number of clicks and the number of pushes corresponding to a characteristic threshold v of each characteristic section, p 1 is a ratio of the number of clicks to the number of pushes corresponding to the leaf node S, and p 2 is a ratio of the number of clicks to the number of pushes corresponding to S v .
This invention relates to a method for selecting data content to be pushed to a terminal based on user attributes. The method addresses the challenge of efficiently determining which content to push to users by analyzing user behavior data to maximize engagement. The system evaluates candidate user attribute types, such as demographics or preferences, to identify the most relevant attributes for content selection. For each attribute type, the method calculates an information gain metric to assess its predictive power in determining user engagement. The information gain is derived from entropy calculations, which measure the uncertainty in user clicks relative to content pushes. Specifically, the method computes the entropy of a leaf node representing a user attribute type and compares it to the entropy of subsets divided by characteristic thresholds within that attribute. The entropy is calculated using the ratio of clicks to pushes, where higher information gain indicates a more discriminative attribute for content selection. This approach optimizes content delivery by prioritizing attributes that best predict user engagement, improving the efficiency of push notifications.
6. The method for selecting data content to be pushed to a terminal according to claim 1 , wherein searching for the decision tree object corresponding to the data content further comprises: creating a decision tree object corresponding to the data content in a case that the decision tree object corresponding to the data content is not found, wherein a root node of the created decision tree object is a leaf node; and allocating a default selection reference value for the data content.
This invention relates to data content delivery systems, specifically methods for selecting and pushing data content to terminals based on decision tree objects. The problem addressed is efficiently managing and delivering relevant data content to terminals, particularly when decision tree objects for the content do not already exist. The method involves searching for a decision tree object corresponding to the data content. If no such object is found, a new decision tree object is created. The root node of this new object is designated as a leaf node, simplifying the structure for initial use. Additionally, a default selection reference value is allocated to the data content, providing a baseline for content selection decisions. This approach ensures that even previously unprocessed data content can be integrated into the system with minimal setup, allowing for dynamic and scalable content delivery. The decision tree object creation process enables the system to handle new or unclassified data content without prior configuration, improving adaptability. The default selection reference value serves as a fallback mechanism, ensuring content can be evaluated and pushed to terminals even when specific decision criteria are not yet established. This method enhances the efficiency and flexibility of content delivery systems, particularly in scenarios where content diversity or user preferences require dynamic adjustments.
7. The method for selecting data content to be pushed to a terminal according to claim 2 , wherein searching for the decision tree object corresponding to the data content further comprises: creating a decision tree object corresponding to the data content in a case that the decision tree object corresponding to the data content is not found, wherein a root node of the created decision tree object is a leaf node; and allocating a default selection reference value for the data content.
This invention relates to a method for selecting and pushing data content to a terminal device, specifically addressing the challenge of efficiently managing and distributing content in systems where decision-making logic must be dynamically generated or retrieved. The method involves searching for a decision tree object associated with specific data content to determine how the content should be selected and pushed to a terminal. If no existing decision tree object is found for the data content, the system automatically creates a new decision tree object. The root node of this newly created decision tree is designated as a leaf node, simplifying the structure for initial use. Additionally, a default selection reference value is assigned to the data content, ensuring that a baseline decision criterion is available even when no prior decision logic exists. This approach enables adaptive content distribution by dynamically generating decision structures when necessary, improving flexibility and scalability in content management systems. The method ensures that content selection remains consistent and efficient, even in scenarios where decision trees must be created on demand.
8. The method for selecting data content to be pushed to a terminal according to claim 3 , wherein searching for the decision tree object corresponding to the data content further comprises: creating a decision tree object corresponding to the data content in a case that the decision tree object corresponding to the data content is not found, wherein a root node of the created decision tree object is a leaf node; and allocating a default selection reference value for the data content.
This invention relates to a method for selecting and pushing data content to a terminal device, particularly in systems where content is dynamically chosen based on decision trees. The problem addressed is the efficient and adaptive selection of data content for transmission to terminals, ensuring relevant content is delivered while minimizing unnecessary data transfer. The method involves searching for a decision tree object associated with specific data content. If no such decision tree object exists, the system automatically creates one. The newly created decision tree object has a root node that functions as a leaf node, simplifying the structure for initial content selection. Additionally, a default selection reference value is assigned to the data content, which serves as a baseline for determining its relevance or priority in subsequent selections. The decision tree object is used to evaluate and select data content for pushing to a terminal. The root node acting as a leaf node allows for immediate content evaluation without requiring a complex hierarchical structure. The default selection reference value ensures that even newly created decision tree objects have a defined starting point for content selection decisions. This approach improves the efficiency of content delivery by dynamically generating decision structures when needed and providing a standardized reference value for content prioritization. The method is particularly useful in systems where content relevance or user preferences are not pre-defined, requiring adaptive selection mechanisms.
9. The method for selecting data content to be pushed to a terminal according to claim 1 , wherein generating the selection reference value based on the number of clicks and the number of pushes further comprises: acquiring a pricing weighting coefficient corresponding to the data content, and multiplying the pricing weighting coefficient with a ratio of the number of clicks to the number of pushes, to obtain a selection reference value for the data content.
This invention relates to a method for selecting data content to be pushed to a terminal device, addressing the challenge of efficiently determining which content to distribute based on user engagement metrics. The method involves generating a selection reference value for data content by analyzing user interactions, specifically the number of clicks and the number of pushes. To refine this selection process, the method further incorporates a pricing weighting coefficient, which is multiplied by the ratio of clicks to pushes. This weighted calculation produces a selection reference value that prioritizes content based on both user engagement and pricing factors, ensuring that the most relevant and commercially viable content is pushed to terminals. The method dynamically adjusts content selection by considering both user behavior and financial incentives, optimizing content distribution for both user satisfaction and business objectives. This approach enhances the efficiency of content delivery systems by balancing engagement metrics with pricing considerations, ensuring that the most impactful content is prioritized.
10. The method for selecting data content to be pushed to a terminal according to claim 1 , wherein acquiring the data content further comprises: prefiltering data content by keyword matching based on the characteristic value, corresponding to the user identifier, in the preset user attribute type.
This invention relates to a method for selecting and pushing data content to a terminal based on user attributes. The problem addressed is efficiently delivering relevant data to users by filtering content according to predefined user characteristics. The method involves acquiring data content and selecting specific content to push to a terminal associated with a user. The selection process includes prefiltering the data content by matching keywords in the content to characteristic values associated with the user's identifier. These characteristic values correspond to preset user attribute types, such as demographics, interests, or behavior patterns. By prefiltering content in this way, the system ensures that only relevant data is considered for further processing and delivery, improving efficiency and personalization. The method may also involve additional steps such as determining the user's current state or context, analyzing the data content's relevance, and dynamically adjusting the selection criteria based on real-time factors. The goal is to optimize content delivery by reducing unnecessary processing and ensuring that the pushed content aligns with the user's attributes and preferences. This approach enhances user engagement and satisfaction by providing more personalized and timely information.
11. A device for selecting data content to be pushed to a terminal, the device comprising a processor and a memory for storing program instructions, wherein the processor is configured to execute the program instructions to: acquire a user identifier, and acquire a characteristic value, corresponding to the user identifier, in a preset user attribute type; acquire data content, and search for a decision tree object corresponding to the data content, wherein a tree node of the decision tree object comprises a branch node and a leaf node, the branch node has a one-to-one correspondence with the user attribute type, and a characteristic threshold of each characteristic section in the user attribute type is stored in the branch node corresponding to the user attribute type, a sub node of the branch node has a one-to-one correspondence with the characteristic threshold, and the number of clicks and the number of pushes corresponding to a characteristic threshold corresponding to the leaf node are stored in the leaf node; locate a leaf node corresponding to the user identifier in the decision tree object based on the characteristic value, corresponding to the user identifier, in the preset user attribute type, wherein the characteristic value matches with a characteristic threshold corresponding to each tree node on a path from a root node of the decision tree object to the located leaf node; and acquire the number of clicks and the number of pushes stored in the located leaf node, generate a selection reference value based on the number of clicks and the number of pushes, and select, based on the selection reference value, data content to be pushed to a terminal corresponding to the user identifier.
This invention relates to a system for selecting and pushing data content to user terminals based on user attributes and historical interaction data. The system addresses the challenge of efficiently delivering relevant content to users by leveraging decision trees that encode user behavior patterns. The device includes a processor and memory storing instructions to execute the following steps. First, it acquires a user identifier and retrieves a corresponding characteristic value from a predefined user attribute type, such as age, location, or browsing history. Next, it retrieves data content and searches for an associated decision tree object. The decision tree consists of branch nodes and leaf nodes. Each branch node corresponds to a user attribute type and contains characteristic thresholds that segment the attribute into ranges. Sub-nodes of a branch node correspond to these thresholds. Leaf nodes store historical interaction data, including the number of clicks and pushes for each threshold range. The system then locates the appropriate leaf node for the user by traversing the decision tree, matching the user's characteristic value against thresholds in each branch node along the path. Once the correct leaf node is found, the system calculates a selection reference value based on the stored click and push counts. This value determines whether the data content should be pushed to the user's terminal. The approach optimizes content delivery by dynamically adjusting selections based on user-specific behavior patterns encoded in the decision tree structure.
12. The device for selecting data content to be pushed to a terminal according to claim 11 , wherein the processor is further configured to execute the program instructions to: receive an uploaded browsing history, and acquire a user identifier corresponding to the uploaded browsing history and data content corresponding to the uploaded browsing history; and acquire a decision tree object corresponding to the data content, acquire a characteristic value, corresponding to the user identifier, in the preset user attribute type, locate a leaf node corresponding to the user identifier in the decision tree object based on the acquired characteristic value, and increase the number of clicks and the number of pushes stored in the located leaf node based on the uploaded browsing history.
This invention relates to a system for selecting and pushing data content to user terminals based on browsing history and user attributes. The problem addressed is the need for personalized content delivery by analyzing user behavior and preferences to improve relevance and engagement. The device includes a processor that executes program instructions to receive a user's browsing history, which is then used to identify the user and the data content they interacted with. The system retrieves a decision tree object associated with the data content, which categorizes users based on predefined attribute types. The user's attributes are matched against the decision tree to locate a specific leaf node, which tracks metrics such as the number of clicks and pushes for that content-user segment. These metrics are updated based on the browsing history, allowing the system to refine content recommendations over time. The decision tree structure enables efficient segmentation of users and content, optimizing push notifications for higher engagement. The system dynamically adjusts content delivery by analyzing historical interactions and user profiles, ensuring tailored recommendations.
13. The device for selecting data content to be pushed to a terminal according to claim 12 , wherein the processor is further configured to execute the program instructions to: acquire the number of clicks and the number of pushes, corresponding to the data content, in the uploaded browsing history; and acquire a branch node on the path from the root node to the located leaf node in the decision tree object, acquire a preset candidate user attribute type besides a user attribute type corresponding to the branch node on the path, and add, by category, the number of clicks and the number of pushes, corresponding to the data content, acquired in the uploaded browsing history based on each characteristic section in the candidate user attribute type.
This invention relates to a system for selecting and pushing data content to user terminals based on user behavior and attributes. The problem addressed is the need for more accurate and personalized content delivery by analyzing user interactions and attributes to improve content recommendation efficiency. The device includes a processor that executes program instructions to analyze browsing history data. Specifically, it acquires the number of clicks and pushes associated with specific data content from the browsing history. The system uses a decision tree object to locate a leaf node corresponding to the data content and then identifies branch nodes along the path from the root node to this leaf node. For each branch node, the processor retrieves a preset candidate user attribute type that differs from the attribute type associated with the branch node. The system then categorizes and aggregates the number of clicks and pushes for the data content based on each characteristic section within the candidate user attribute type. This allows the system to refine content selection by considering multiple user attributes beyond those directly linked to the decision tree path, enhancing the relevance of pushed content to individual users. The approach improves content recommendation accuracy by leveraging both explicit user interactions and broader attribute-based segmentation.
14. The device for selecting data content to be pushed to a terminal according to claim 13 , wherein the processor is further configured to execute the program instructions to: generate an information gain corresponding to the candidate user attribute type based on the number of clicks and the number of pushes, corresponding to each characteristic section in the candidate user attribute type, stored in the located leaf node by category; search for a candidate user attribute type having an information gain, a difference between which and an information gain of other searched user attribute type is greater than or equal to an information gain threshold; and set the located leaf node as a branch node and generate a leaf node of the branch node based on a characteristic threshold of a characteristic section in the found candidate user attribute type, in a case that the candidate user attribute type is found.
This invention relates to a system for optimizing data content delivery to user terminals by dynamically selecting relevant user attributes to improve content targeting. The system addresses the challenge of efficiently identifying the most informative user attributes to enhance the accuracy of content recommendations, thereby improving user engagement and system performance. The device includes a processor that executes program instructions to analyze user attribute types and their impact on content selection. Specifically, the processor calculates an information gain for each candidate user attribute type based on user interactions, such as clicks and content pushes, stored in a decision tree structure. The information gain quantifies the predictive value of each attribute type by comparing the number of clicks and pushes across different characteristic sections within the attribute type. The processor then searches for candidate user attribute types where the difference in information gain compared to other attribute types meets or exceeds a predefined threshold. If such an attribute type is found, the system updates the decision tree by converting the current leaf node into a branch node and generating a new leaf node based on a characteristic threshold derived from the selected attribute type. This adaptive approach refines the decision tree structure to prioritize the most relevant user attributes, ensuring more accurate and efficient content delivery. The system dynamically adjusts the tree structure to optimize content targeting over time.
15. The device for selecting data content to be pushed to a terminal according to claim 14 , wherein the processor is further configured to execute the program instructions to: calculate an information gain of a user attribute type A of a leaf node S according to: { G ( A ) = Entropy ( S ) - ∑ v ∈ F A p ( v ) Entropy ( S v ) Entropy ( S ) = - p 1 log 2 ( p 1 ) - ( 1 - p 1 ) log 2 ( 1 - p 1 ) Entropy ( S v ) = - p 2 log 2 ( p 2 ) - ( 1 - p 2 ) log 2 ( 1 - p 2 ) wherein F A is a set of characteristic sections in the user attribute type A, v is a characteristic threshold of each characteristic section in the user attribute type A, p(v) is a distribution probability of the number of pushes in each characteristic section in the user attribute type A, S v is a set of the number of clicks and the number of pushes corresponding to a characteristic threshold v of each characteristic section, p 1 is a ratio of the number of clicks to the number of pushes corresponding to the leaf node S, and p 2 is a ratio of the number of clicks to the number of pushes corresponding to S v .
This invention relates to a system for selecting and pushing data content to user terminals based on user attributes and interaction metrics. The system addresses the challenge of efficiently delivering relevant content to users by analyzing user behavior and optimizing content selection using information gain calculations. The device includes a processor that executes instructions to evaluate user attributes and interaction data, such as clicks and pushes, to determine the most effective content to deliver. The processor calculates the information gain of a user attribute type by comparing the entropy of a leaf node (representing a subset of users) before and after partitioning the data based on characteristic thresholds. Entropy is computed as a measure of uncertainty, where lower entropy indicates more predictable user behavior. The system partitions user data into sections based on attribute thresholds and evaluates the distribution probability of content pushes and clicks within each section. By comparing the entropy of the original data set to the weighted entropy of the partitioned sections, the system identifies the most informative user attributes for content selection. This approach improves content relevance and user engagement by leveraging statistical measures to optimize push notifications.
16. The device for selecting data content to be pushed to a terminal according to claim 11 , wherein the processor is further configured to execute the program instructions to: create a decision tree object corresponding to the data content in a case that the decision tree object corresponding to the data content is not found, wherein a root node of the created decision tree object is a leaf node; and allocate a default selection reference value for the data content.
This invention relates to a system for selecting and pushing data content to a terminal device, addressing the challenge of efficiently managing and delivering relevant content to users. The system includes a processor that executes program instructions to create and manage decision tree objects for data content. When a decision tree object for specific data content is not found, the processor generates a new decision tree object where the root node is also a leaf node. This simplifies the decision-making process by ensuring that the initial state of the decision tree is straightforward and easily modifiable. Additionally, the processor assigns a default selection reference value to the data content, which serves as a baseline for subsequent content selection decisions. The decision tree structure allows for hierarchical organization of content selection criteria, enabling dynamic updates and refinements based on user interactions or other factors. This approach improves the efficiency and accuracy of content delivery by providing a structured framework for decision-making, reducing the computational overhead associated with complex content selection algorithms. The system is particularly useful in environments where real-time content delivery is required, such as in digital advertising, personalized recommendations, or automated content distribution.
17. The device for selecting data content to be pushed to a terminal according to claim 12 , wherein the processor is further configured to execute the program instructions to: create a decision tree object corresponding to the data content in a case that the decision tree object corresponding to the data content is not found, wherein a root node of the created decision tree object is a leaf node; and allocate a default selection reference value for the data content.
This invention relates to a system for selecting and pushing data content to a terminal device, addressing the challenge of efficiently managing and delivering content based on user preferences or system criteria. The device includes a processor that executes program instructions to handle data content selection dynamically. When a decision tree object for specific data content is not found, the processor creates a new decision tree object with a root node that functions as a leaf node. This structure simplifies the decision-making process for content selection. Additionally, the processor assigns a default selection reference value to the data content, ensuring that even new or unclassified content can be processed without requiring prior categorization. The decision tree object allows for hierarchical organization of content, enabling efficient filtering and prioritization based on predefined rules or user behavior. The default selection reference value serves as a baseline for content evaluation, which can be refined over time as more data is collected. This approach improves content delivery efficiency by reducing the need for manual intervention and ensuring consistent handling of both new and existing data.
18. The device for selecting data content to be pushed to a terminal according to claim 11 , wherein the processor is further configured to execute the program instructions to: acquire a pricing weighting coefficient corresponding to the data content, and multiply the pricing weighting coefficient with a ratio of the number of clicks to the number of pushes, to obtain a selection reference value for the data content.
This invention relates to a system for optimizing data content delivery to user terminals based on user engagement metrics. The problem addressed is the inefficient selection of content for push notifications, where content is often pushed without considering user interaction patterns, leading to low engagement and wasted resources. The device includes a processor that executes program instructions to analyze user interactions with pushed content. Specifically, it acquires a pricing weighting coefficient for each piece of data content, which represents the relative importance or value of the content. The processor then calculates a selection reference value by multiplying this coefficient with the ratio of the number of user clicks to the number of times the content was pushed. This ratio indicates how frequently users engage with the content when it is delivered. The selection reference value helps prioritize content that has historically generated higher engagement, ensuring that more valuable or interactive content is more likely to be pushed to users in the future. The system improves content delivery efficiency by dynamically adjusting push decisions based on real-world user behavior, increasing the likelihood of successful engagement and reducing the push of low-interaction content. The processor may also be configured to perform additional tasks, such as monitoring user preferences or adjusting push strategies in real-time. The overall goal is to enhance user experience and optimize resource allocation in content distribution systems.
19. The device for selecting data content to be pushed to a terminal according to claim 11 , wherein the processor is further configured to execute the program instructions to: prefilter data content by keyword matching based on the characteristic value, corresponding to the user identifier, in the preset user attribute type.
This invention relates to a system for selecting and pushing data content to a terminal based on user attributes. The problem addressed is efficiently delivering relevant content to users by filtering data according to predefined user characteristics. The device includes a processor that executes program instructions to prefilter data content using keyword matching. The filtering is based on a characteristic value associated with a user identifier, which corresponds to a preset user attribute type. The system ensures that only content matching the user's attributes is selected for delivery, improving content relevance and user engagement. The processor may also perform additional filtering steps, such as semantic analysis or user behavior analysis, to further refine the content selection. The device is designed to operate in environments where large volumes of data are generated, and users have diverse interests, requiring precise targeting to enhance user experience. The invention optimizes content delivery by reducing irrelevant data transmission, conserving network resources, and increasing user satisfaction. The system is particularly useful in applications like personalized news feeds, advertising, or recommendation engines where tailored content delivery is critical.
20. A non-transitory computer storage medium comprising computer executable instructions that, when performed by a processor, cause the processor to perform a method for selecting data content to be pushed to a terminal, wherein the method for selecting data content to be pushed to a terminal comprises: acquiring a user identifier, and acquiring a characteristic value, corresponding to the user identifier, in a preset user attribute type; acquiring data content, and searching for a decision tree object corresponding to the data content, wherein a tree node of the decision tree object comprises a branch node and a leaf node, the branch node has a one-to-one correspondence with the user attribute type, and a characteristic threshold of each characteristic section in the user attribute type is stored in the branch node corresponding to the user attribute type, a sub node of the branch node has a one-to-one correspondence with the characteristic threshold, and the number of clicks and the number of pushes corresponding to a characteristic threshold corresponding to the leaf node are stored in the leaf node; locating a leaf node corresponding to the user identifier in the decision tree object based on the characteristic value, corresponding to the user identifier, in the preset user attribute type, wherein the characteristic value matches with a characteristic threshold corresponding to each tree node on a path from a root node of the decision tree object to the located leaf node; and acquiring the number of clicks and the number of pushes stored in the located leaf node, generating a selection reference value based on the number of clicks and the number of pushes, and selecting, based on the selection reference value, data content to be pushed to a terminal corresponding to the user identifier.
A system for selecting and pushing data content to a terminal based on user attributes and historical interaction data. The system addresses the challenge of efficiently delivering relevant content to users by leveraging decision trees to analyze user characteristics and past engagement metrics. The method involves acquiring a user identifier and retrieving a corresponding characteristic value from a predefined user attribute type, such as demographics or behavior patterns. Data content is then analyzed by searching for an associated decision tree object, which contains branch nodes and leaf nodes. Each branch node corresponds to a user attribute type and stores characteristic thresholds for that attribute. Sub-nodes under each branch node correspond to these thresholds. Leaf nodes store historical data, including the number of clicks and pushes for each characteristic threshold. The system locates the appropriate leaf node by matching the user's characteristic value against the thresholds in the decision tree, following a path from the root node to the leaf node. Once the leaf node is identified, the system calculates a selection reference value based on the stored click and push counts, then uses this value to select the most suitable data content to push to the user's terminal. This approach optimizes content delivery by dynamically adjusting recommendations based on user attributes and engagement history.
Unknown
September 29, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.